ACHIEVING EFFICIENCY IN ABRASIVE BLAST

4Fig. 4: Configuration and dimension of Nozzle B estimated by X-ray image analysisTable 2: Configuration and Dimension of Tested NozzlesNozzleLength (mm)Throat 15011.52.19.3*DE215.9 185.79.5111.21.38.57.9*Intermittent dischargingTable 3: Abrasive Flow Rate, Nozzle Pressureand Air Consumption with Nozzle Type Valve OpeningNozzleA (11.5Ø, 150 mm)B (11Ø, 120 mm)C (12.5Ø, 117 mm)D (9.5Ø, 185.7 mm)E (11Ø, 215.9 mm)Abrasivevalve # .955.935.85.755.73Air 40330310295*Test conditions: hopper pressure (6.6 kg/cm2); kgf/cm2x14.2 psi; grit (avg. 1 mm); hose I.D. (32Ø)Table 4: Evaluation of Repeatabilityof the Special Abrasive Metering Valves# of turns3.54.55.56.5(Unit: kg/30 sec)#1 valve#2 valve19.2618.7616.8616.6411.6611.286.446.0Step 2: Nozzle Type vs. Blasting Pattern (Width)In the evaluation of the blasting pattern, especially the blastingwidth, the test focused on the three kinds of blasting nozzles thatperformed best in step 1. In the case of identical blasting performance, the blaster usually prefers the nozzle that can create thelarger blasting width.Therefore, blasting pattern is also a critical factor in selectingthe optimum blasting nozzle. Figure 3 shows the blasting patternswith selected blasting nozzles with the same distance (750 mm)between the blasting nozzle and the target surface for all tests.The distance was maintained by using a small rod attached to theblasting hose that extended to the target surface. Nozzle E wasinferior to nozzle B and nozzle A in terms of blasting width (in mm):Nozzle C (156) Nozzle B (126) Nozzle A (122) Nozzle E Nozzle D (118).Step 3: Nozzle Configuration vs. Abrasive Flow PatternNozzle design also has a great effect on blasting productivity. Therefore, a number of researchers have tried to improve the nozzlegeometry. As a result, the Venturi nozzle is commonly used becauseof its higher blasting efficiency compared with that of a conventionalnozzle.X-ray images were taken from the cross-section of the blastingnozzles evaluated to identify their internal configurations. Figure 4and Table 2 show the configuration and dimension of Nozzle B, including the total length, throat, and converging and diverging section.The shape of the diverging section is responsible for the speed ofthe abrasive particles at the nozzle exit and for the blasting width.The discharge stability of the abrasive stream is due to the converging section.The rank of diverging angles derived from the X-ray images revealed the same blasting pattern tendency seen in step 2. Nozzle Bshowed continuous discharge characteristics regardless of abrasivesize, while a significant problem was encountered with Nozzle A.Nozzle A exhibited intermittently discharging abrasive at the nozzleexit when the abrasive size was less than 0.5 mm. These resultsmay be attributed to the relatively rapid converging angle (9.3 degrees) of Nozzle A against that (7.8 degrees) of Nozzle B (Fig. 5 andTable 2). That is, it can be postulated that the possibility of a bottleneck in a blasting nozzle is increased with an increase in its converging angle and a decrease in abrasive size. These results indicatethat the use of X-ray imaging to analyze nozzle configuration can bevaluable for predicting the working characteristics of a blasting nozzle.Step 4: Nozzle Type vs. Air ConsumptionAir consumption. an additional factor in selecting the optimum blasting nozzle, isclosely related to the cost of abrasive blasting. Abrasive flow rate, nozzle pressure,and air consumption for each blasting nozzle along with abrasive valve openingare summarized at Table 3. At a similar abrasive flow rate ( 1,300 kg/hr), the airconsumption of Nozzle B was relatively lower than that of Nozzle A. Based on theresults in Table 3, Nozzle B proved to be more suitable than the other nozzles foractual field blasting conditions in the shipbuilding industry.

5Fig. 5: Older spring-type of abrasive metering valve stillused in some shipyardsAbrasive Metering ValveAn abrasive control valve that can precisely meter abrasive media under variable conditions should be selected and adjusted to optimize the cleaning rate.Too little abrasive introduced into the airstream results in an incompletelyfilled blast pattern, which slows production and leaves areas on the substrateuntouched. Too much abrasive causes abrasive particle collision, whichwastes energy and disperses particles unevenly within the blast pattern. Aproperly adjusted metering valve ensures that the maximum amount of cleaning is gained from each abrasive particle.5 Figure 5 shows the spring type ofabrasive metering valve still used in some shipyards. However, this older typeof valve lacks repeatability in metering abrasive flow rate and needs much effort to calibrate and select the proper valve turns.A new type of abrasive metering valve, which provided precise control andrepeatability in abrasive flow rate, markedly improved control of abrasive flow(Fig. 6). Abrasive flow rates of two special abrasive metering valves arenearly identical, indicating that the new type of valve is precise and can beused to effectively control the qualitity of abrasive media (Table 4) with goodrepeatability.Abrasive Size vs. Blasting ProductivityIt is well known that the finer the abrasive is, the faster the surface cleaningrate is. Settles et al. (1995) reported that large abrasive particles could not befully accelerated at the nozzle exit because their inertia generally preventedthem from accelerating as rapidly as the small abrasive particles.5 On theother hand, some abrasive manufacturers insisted that the operating mixFig. 6: Newer type of abrasive metering valve that allowsmore precision than spring type(work mix) of abrasives of various sizes is a vital factor in improving blastingproductivity. Therefore, the effect of an abrasive manufacturer’s suggestedoperating mix on cleaning rate was evaluated, as was the effect of abrasive size.Blasting productivity as a function of representative abrasive sizes was measured at the following experimental conditions: Nozzle A and hopper pressure (6.6 kgf/cm2). The cleaning ratewas inversely proportional to the abrasive particle size. The highest cleaning rate wasachieved with the smaller abrasive particles because they allowed a higher number of impactson a unit area and a resulting high peak count (except for abrasive of 0.3 mm, which was in theexcessive discharging range).On the other hand, the effect of operating mix on blasting efficiency did not meet our expectations. For secondary surface treatment of a shop-primed surface, less and looser contamination was expected. It can be postulated that even though they impart less impact energythan large abrasives, small abrasive particles probably have sufficient impact energy for complete removal of these contaminants. Nozzle A exhibited the phenomenon of intermittently discharging abrasive at the nozzle exit when the abrasive size was less than 0.5 mm. Reducingthe abrasive size from 1 mm to 0.5 mm increased the blasting productivity by 40%. (See Table 5.)Table 5: Comparison of Blasting Productivity of Various Sized AbrasivesAbrasive size (avg.)Blasting productivity (m2/hr)Relative efficiency (%)118.7100*discharge excessive; **pressure set at 6.6 8RemarksNozzle AHopper **

6Abrasive Size vs. Surface ProfileLarge abrasives cut deeper and produce deeper proftles than smaller particles with thesame composition and shape. However, Keane et al. (1976) reported that coarser abrasivestend to lead to more hackles (large surface deformations) than do the small abrasives.7Hackles are often considered the ‘onset point of pinpoint rust’ because a hackle tip will protrude through the thin layer of coating, allowing premature coating failure. Moreover, theIMO’s new provisions for coating ballast tanks require maintaining the surface roughness ofa blast-cleaned surface between 30 and 75 µm after secondary surface preparation. Thus,the profile height produced by variously sized abrasives has received considerable attentionin new shipyards all over the world.The surface profile after each blasting test was measured inTable 6: Surface Profile with Various Sizedaccordance with ISO 8503-4 (“Method for Calibration for ISOAbrasives (hopper pressure: 6.6 kgf/cm2)Surface Profile Comparators and the Determination of SurfaceProfile”) with a proprietary stylus instrument. Among the abrasivesAbrasiveRmax(µm)Rz(µm)evaluated, those measuring 0.5 mm and below met the IMOBare Steel8.0 11.96.5 7.3PSPC’s acceptance criteria, whereas the others didn’t (Table 6).1 mm97.3 106.2The 0.3 mm abrasives, however, produced a surface texture0.7 mm96.1 111.780.2 94.5quite different from those produced by the other abrasives andO/M90.3 101.478.6 84.4generated more dust during the blasting, causing poor visibility0.5 mm79.2 96.664 73.9for workers. From these results, we concluded that 0.5 mm0.3 mm44.6 69.238.2 48abrasive is the most suitable abrasive for blast cleaning.Abrasive Hardness vs. Blasting ProductivityThe hardness of abrasive affects the cleaning and breakdown rates. A general rule for surfacepreparation has been that hard abrasives generate deeper and faster cutting action than softabrasives. For evaluating the effect of abrasive hardness on cleaning rates, we tested andcompared the performance of conventional and harder abrasives. Hardness of the conventional abrasive averaged 56 HRC, whereas harder abrasive, identified here as SHG (SpecialHigh Grit) averaged 64 HRC. With the use of 0.5 mm virgin abrasives, the SHG increased thecleaning rate by as much as 10%. For actual field application of SHG, however, further detailedtesting on the lifetime of SHG is necessary because hard abrasive is generally known to havea higher propensity to fracture (friability) than conventional abrasive.However, in terms of profile depth or height, the surface profile created by the SHG duringperformance testing was 65–78 µm, close to the surface roughness of 64–74 µm created withconventional abrasive (56 HRC). This unexpected result was probably caused by the shorterresidence time of the harder abrasive on a unit area during blasting, compared to the residence time of the abrasive of conventional hardness. That is, if the conventional and harderabrasives were to impact an equal area for the same amount of time, the harder abrasive mayproduce a higher surface profile than that of a conventional abrasive.However, during actual blasting in the shop, shipyard, or field, when surface preparation onone area is completed, the blaster rapidly begins to clean the next unblasted surface area.Hence, the harder abrasive has a short residence time on the surface (per unit area) comparedto that of a conventional abrasive. Because of its shorter residence time, a harder abrasivemight not be able to exhibit its full potential to create a deeper profile than that of an abrasive ofconventional hardness.Nozzle Pressure vs. Blasting ProductivitySettles et al. (1995) reported that increasing the nozzle pressure increased the dynamicpressure (1 2 rU2) proportionally, which, in turn, proportionally increased the drag force that accelerates the particle through the blasting nozzle.6 (r is the fluid density in kg/m3 and U is thefluid velocity in m/sec.) In addition, an abrasive equipment manufacturer qualitatively expressed the effect of nozzle pressure on blasting productivity as follows: “a reduction of 1 psireduces the productivity by 1.5%.”8

7To identify the correlation between nozzle pressure and blasting productivity, tests were conducted by varying the nozzle type and pressure. Test results with Nozzles A and B showed thatblasting productivity increased as hopper pressure increased, although Nozzle B showed relatively lower performance against Nozzle A. The abrasive flow rate of Nozzle B was lower thanthat of Nozzle A with the valve opening at 6 turns (Table 7). So, despite its higher pressure,Nozzle B may be inferior to Nozzle A in blasting productivity.Table 7: Predicted and Measured Blasting Productivity with Nozzle Pressure*Hopper pressure (kgf/cm2)Nozzle pressure (kgf/cm2)Abrasive flow rate (kg/hr)Air consumption (Nm3/hr)Blasting Productivity (m2/hr)Relative efficiency (%)Nozzle A (6.6)5.61,45033026.38100Nozzle A (7.0)5.951,50037028.08106.5Nozzle B (7.0)5.981,31531027.27103.4*Valve opening- 6 turns; abrasive size: 0.5 mm; kgf/cm2x14.2 psiNozzle pressure increased by 0.35 kgf/cm2 when the hopper pressure increased from 6.6kgf/cm2 to 7.0 kgf/cm2. This slight pressure difference between nozzle and hopper was explained by a pressure drop caused by the abrasive particles’ interference in the blasting hoseand on the internal surface during the pneumatic transport of the abrasive/air mixture. The relative blasting productivity of Nozzle A was increased up to 106% by an increase in the nozzlepressure. From the above results, we also determined that the empirical equation proposed bythe equipment manufacturer is reasonably accurate and can be used (although with a minorlimitation) for predicting the influence of nozzle pressure on blasting productivity.On the other hand, after blasting performance was tested with an increase in nozzle pressure, surface roughness or profile was 70–81 µm, which was higher than profiles (64–74 µm)obtained at lower nozzle pressure (6.6 kgf/cm2). These results have two important implications.First, unlike abrasive hardness, nozzle pressure has a great effect on surface profile. Second,increasing the nozzle pressure also increases the possibility of exceeding the IMO’s PSPC requirement (30–75 µm) for surface roughness. However, according to Baek et al. (1995), thisserious problem could be solved by adopting sweep blasting (SSPC-SP 7, Brush-Off BlastCleaning) with finer-sized blasting abrasive instead of performing full blasting.9 Baek et al. reported that after full blasting, the surface roughness of a mock-up block averaged 83.1 µm atthe field conditions with abrasive size averaging 0.75 mm and nozzle pressure at 6 kgf/cm2).In contrast, surface roughness was decreased to an average of 75.2 µm when sweep blastingwas used.In our experience, areas with looser contamination, such as areas contaminated with zincsalt, appeared mainly on the shop-primed surface during the secondary surface treatment.Accordingly, it might seem that the sweep blasting could remove such contamination satisfactorily and efficiently. Based on the above results, we judged that blasting method, nozzlepressure, and abrasive size were far more important than abrasive hardness in terms ofblasting productivity and surface roughness.ConclusionsFirst, for higher blasting productivity with less abrasive consumption, selecting the optimalabrasive-to-air mixing ratio is important. The ratio should be based on the blasting performancecurve, which shows that blasting productivity tends to gradually increase with the abrasivefeeding rate, up to a critical value, and then productivity is maintained constantly. In addition, toachieve the optimum abrasive feeding rate to the actual blasting process, the proper choiceand adjustment of an abrasive metering valve with high repeatability are essential.

8Second, the configuration of the blasting nozzle governs work efficiency and air consumptionas well as blasting productivity and its pattern. Accordingly, the selection of an optimum nozzleshould be based on a comprehensive evaluation of these blasting-related parameters ratherthan a simple evaluation of cleaning rate. It is possible to predict the blasting pattern and working characteristics of a blasting nozzle by evaluating its diverging and converging angle usingthe newly proposed X-ray image analysis method.Third, by quantitatively assessing the influence of abrasive size, hardness, and nozzle pressure on blasting performance, we have shown three factors that significantly affect blastingproductivity. The proper combination of abrasive-to-air mix ratio, abrasive metering valve, andnozzle can dramatically increase the blasting productivity, whereas the effect of abrasive hardness on surface profile is unexpectedly insignificant. Furthermore, our study shows that usingsweep blasting as much as possible, with small abrasive, and controlling the nozzle pressureproperly can help shipyards meet the IMO’s recently adopted PSPC provisions without reducing blasting produclivity.References1. C. G. Munger, Corrosion Protection by Protective Coatings, Second edition, NACE, 1999.2. IMO, “Performance Standards for Protective Coatings,” Sub-Committee on Ship Designand Equipment (DE), 49th Session, 2006.3. J. LeCompte and G. Mort, “The Importance of Proper Blasting Nozzle Selection and Use,”JPCL, November 1986, pp. 3-4.4. W. S. Holt et al., “How Nozzle Pressure and Feed Rate Affect the Productivity of Dry Abrasive Blasting,” JPCL, October 2001, pp. 82-104.5. KTA-Tator, Inc., “Evaluation of Substitute Materials for Silica Sand in Abrasive Blasting,”September 1998.6. G. S. Settles et al., “A Scientific View of the Productivity of Abrasive Blasting Nozzles,”JPCL, April 1995, pp. 28-102.7. J. D. Keane et al., “Surface Profile for Anti-Corrosion Paints,” SSPC, 1976, pp. 22-27.8. H. W. Hitzrot, “Reduce Volume of Spent Abrasive in Open Air Blasting,” NSRP Report, 1997.9. K. K. Baek et al., “Effect of Retaining Preconstruction Primer (PCP) on the Quality of HighPerformance Protective Coating Systems,” NACE, Paper No. 05003, 2005.Mr. Han-Jin Bae is a senior research engineer for the ProtectiveCoatings and Corrosion ResearchDepartment of the R&D Center ofHyundai Heavy Industries inCheonha-Dong, Ulsan, Korea. Hehas 6 years of experience in surface preparation technology. He isa registered member of The Korean Society of Industrial & Engineering Chemistry (KSEIC).Dr. Kwang-Ki Baek is the department head of the Protective Coatings and Corrosion ResearchDepartment of the R&D Center ofHyundai Heavy Industries. Hisareas of expertise include protective coatings and linings of steelstructures, corrosion and corrosion control of oil/gas and offshore structures, andmaterials selection and evaluation of materials. Inaddition to serving on the Board of Directors of theCorrosion Science Society of Korea, Dr. Baek is amember of NACE International. He earned hisPh.D in Materials Science and Engineering fromthe Massachusetts Institute of Technology.Mr. Byung-Hun Lee is principal researcher for the Protective Coatingsand Corrosion Research Department ofthe R&D Center of Hyundai HeavyIndustries. He has performed researchon industrial rationalization and theautomation of production equipmentsince 1984. His inventions in paintingand welding equipment have led to more than 30 patents.Mr. Jae-Jin Baek is a senior researchengineer for the Protective Coatingsand Corrosion Research Department ofthe R&D Center of Hyundai Heavy Industries. He has 10 years of experiencein research and development for automatic painting systems and coatingequipment. He is a registered memberof KSEIC.Mr. Ki-Soo Kim is a general manager ofthe block painting department of theshipbuilding division of Hyundai HeavyIndustries. He has engaged in the fieldof corrosion protection (i.e., anticorrosion design, coating inspection, andmanagement of painting) since 1983.He holds a FROSIO certific

Air consumption. an additional factor in selecting the optimum blasting nozzle, is closely related to the cost of abrasive blasting. Abrasive flow rate, nozzle pressure, and air consumption for each blasting nozzle along with abrasive valve opening are summarized at Table 3. At a similar abrasive flow rate ( 1,300 kg/hr), the air